Vaughany's MMA Picks...

Originally Posted by DublinMeUp

Out on a limb, please be nice Nunya =),

Except you're forgetting that nash's equilibrium assumes that all parties play or bet perfectly, meaning to deviate from it would cause you or others to lose EV.
Seeing as typically none of the above mentioned parties will be "playing" perfectly, nash is a sub optimal formula to use in this argument.

Equilibria doesn't mean that each party is making the best decision for themselves, but that they are making the best decision relatively to all other parties actions. In this case, it is in the best interest of the large bettors for everyone to wait until limits increase. It makes no difference to small bettors either way.

This would hold true if we could definite it as a 2x2 matrix. If we could think of Small Bettors as one collective group and Large Bettors as another.

In reality, lines move closer to parity with every dollar bet. If all players collude to wait for the limits to increase, then the one player who does NOT wait gains the most. Each additional player who does NOT wait gains less than the first, but more than the players who are waiting, and quickly we discover that the only equilibria is no players waiting.

The best town is the one in which no one is a thief. The best place to be a thief is a town without one.

I don't have your way with words unfortunately but I'll try,

Nash based his theorem on zero sum games, ie solved games. In it he says that in such a game where each party makes the correct play/action in each given situation it is -EV to deviate from the correct play yourself. However if you or any other participant does deviate, the game is no longer in equilibrium and therefore it is -EV for you and/or others to continue to play as you had. Nash is not about optimal vs sub optimal at all as it only holds true in one situation.

Therefore, you are using nash incorrectly to rationalise your big bettor / small bettor scenario, Since there is never a point where a market is solved it cannot be proven one way or the other that it is in the best interest of either party to take a certain action.

"The best town is the one in which no one is a thief" true and this is nash
"The best place to be a thief is a town without one" also true in literal terms but in equilibrium terms its a contradiction onto itself in that the town stops becoming the best town as soon as you become a thief ie it is no longer in equilibrium.

Originally Posted by Vaughany

Did you buy ttwarrior's shit towel on ebay, Vaughany?

Originally Posted by DublinMeUp

Nash based his theorem on zero sum games,

Quite the contrary - he was the one to expand the scope of game theory beyond zero-sum games!

Originally Posted by DublinMeUp

I don't have your way with words unfortunately but I'll try,

Nash based his theorem on zero sum games, ie solved games. In it he says that in such a game where each party makes the correct play/action in each given situation it is -EV to deviate from the correct play yourself. However if you or any other participant does deviate, the game is no longer in equilibrium and therefore it is -EV for you and/or others to continue to play as you had. Nash is not about optimal vs sub optimal at all as it only holds true in one situation.

Therefore, you are using nash incorrectly to rationalise your big bettor / small bettor scenario, Since there is never a point where a market is solved it cannot be proven one way or the other that it is in the best interest of either party to take a certain action.

"The best town is the one in which no one is a thief" true and this is nash
"The best place to be a thief is a town without one" also true in literal terms but in equilibrium terms its a contradiction onto itself in that the town stops becoming the best town as soon as you become a thief ie it is no longer in equilibrium.

Are you stating that there can be no Nash equilibria in nonzero-sum games? Here is a typical non-zero sum game:



This is easily solvable by assigning probability to Henry's actions if you're Dave. We can move the equilibrium points from square to square by changing the payoffs and probabilities, however, but its solvable at every juncture.

Von Neumann showed that any non-zero sum game becomes a zero sum game when we add in another player (the player whom the losses (or gains) are attributed to), in our model we can include all other bettors or the juice lost to the book itself.

Either way, arguing game theory 101 isn't the point I'm trying to make (and I can't argue any deeper because I never took 102). My original point holds true, wishing and hoping that other players won't hit openers so we can get the good lines doesn't do us any good as the good lines are always going to be hit.

Originally Posted by NunyaBidness

Are you stating that there can be no Nash equilibria in nonzero-sum games?

Yes

Originally Posted by NunyaBidness

We can move the equilibrium points from square to square by changing the payoffs and probabilities, however, but its solvable at every juncture.

I don't think i quite get the game lol, but if your summation is true then it is a solved game, zero sum. All potential actions are known along with all potential repercussions and/or reactions. There are no unknowns. So nash would/could apply.

Originally Posted by NunyaBidness

Von Neumann showed that any non-zero sum game becomes a zero sum game when we add in another player (the player whom the losses (or gains) are attributed to), in our model we can include all other bettors or the juice lost to the book itself.

i'm gonna leave this one, while I don't necessarily agree i don't know enough about Von Neuman to argue either way. Wasn't he the guy to develop the first or one of the first computer architectures? lol

Originally Posted by NunyaBidness

Either way, arguing game theory 101 isn't the point I'm trying to make (and I can't argue any deeper because I never took 102).

Haha same here mate, I'm only giving my point of view.. there is every chance one or both of us have it wrong.

Originally Posted by NunyaBidness

My original point holds true, wishing and hoping that other players won't hit openers so we can get the good lines doesn't do us any good as the good lines are always going to be hit.

If you actually think about this it is a one sided blanket statement that doesn't have any mathematical basis. Who defines what a good line is? We can approximate yes but we cannot say for sure. You are also leaving out the other side of things, many fighters or any other sports prop also drift out in price whether they start favourite or not, the price is changed by the amount of money staked but does that in a vacuum affect the chances of it winning/losing? (by in a vacuum i mean assume all things are equal and that one fighter hasn't had his leg amputated etc)

Its a good example of selective memory, which I think i saw you discuss in regard to poker. People get annoyed when a line is cut but they are happy when odds improve.. which feeling do you think the human brain holds on to in most cases?

Ok lastly, although i enjoyed having a proper discussion rather than the usual flinging of racial or homophobic slurs at each other as seems the norm on here.

Do you agree with this statement;

You can prove nash by using perfect game strategy in a solved game, but you can't prove perfect game strategy by using nash in an unsolved game.

If nash was such an all knowing and self adjusting optimal strategy forecaster there would be no such thing as sports betting or even stock markets, these only exist because they are imperfect. In other words if they like say a game of dice where solved and we could use nash on them they would cease to exist as financial entities.

Anyway I hope you haven't taken this as me calling you out or picking a fight, Because it not that at all. Its more of a two sides of a coin type thing,

Holy f*ck you dorks are gonna scare away all the posters from Vaughn's thread with this math/logic circle jerk... ;p

Vaughn -- any of your crazy books got props for Pierce v Rocha?

Originally Posted by Educ8d Degener8

Holy f*ck you dorks are gonna scare away all the posters from Vaughn's thread with this math/logic circle jerk... ;p

Vaughn -- any of your crazy books got props for Pierce v Rocha?

yeah, take that shit over to my thread.

I appreciate the conversation as well, and am happy to admit when I'm wrong. Although I don't think I am here. It's been many years since Game Theory class, but I do play around with it from time to one (0,1 games in heads up poker simulations and what not).

There are definitely nash equilibria in non-zero sum games.

Originally Posted by DublinMeUp

i'm gonna leave this one, while I don't necessarily agree i don't know enough about Von Neuman to argue either way. Wasn't he the guy to develop the first or one of the first computer architectures? lol

I believe that is true, but he did lots of other stuff as well. I think he is best known for 'Theory of Games and Economic Behavior'. If you think about the statement, that all non-zero sum games can be converted to zero sum games with the addition of a third player, it seems to obviously make sense, although in cases we might have to come up with a strange idea who this player is. Von Neumann was the one who proved it was true, however.

Originally Posted by DublinMeUp

Do you agree with this statement;

You can prove nash by using perfect game strategy in a solved game, but you can't prove perfect game strategy by using nash in an unsolved game.

If nash was such an all knowing and self adjusting optimal strategy forecaster there would be no such thing as sports betting or even stock markets, these only exist because they are imperfect. In other words if they like say a game of dice where solved and we could use nash on them they would cease to exist as financial entities.

I'm not sure what I think of that statement, I'm pretty tired from playing way too much DayZ today, might have to give it another look tomorrow.

Regarding the rest however, I don't know that anyone thinks of Game Theory as being an actual predictor for anything. I think its usefulness comes in modeling problems only. Most matrices are incredibly simplifed. It would be impossible to make actual predictions with it, a la Asimov's Foundation, there are too many actors, too many variables.

The basic problems of game theory have few practical uses, but they make good thought experiments for working through related situations. Infrastructure usage is about the only real practical usage I can think of off the top of my head.

Originally Posted by DublinMeUp

If you actually think about this it is a one sided blanket statement that doesn't have any mathematical basis. Who defines what a good line is? We can approximate yes but we cannot say for sure. You are also leaving out the other side of things, many fighters or any other sports prop also drift out in price whether they start favourite or not, the price is changed by the amount of money staked but does that in a vacuum affect the chances of it winning/losing? (by in a vacuum i mean assume all things are equal and that one fighter hasn't had his leg amputated etc)

I do think there is a mathematical basis there, if you believe in efficient markets. I'm not certain if MMA markets are efficient yet, but if not, they will be within a few years. As far as conditional probability like you mention in your last statement, no the line moving doesn't affect the chances of it winning or losing, but the chances of something winning or losing change the line.

5Dimes clearly believes that MMA markets are efficient, watching their lines bounce and sharpen is fascinating to me. I haven't personally parsed the data, but I am certain that if someone were to bet all vig-free closers over a few thousand events they would be very close to breaking even.

Originally Posted by DublinMeUp

Anyway I hope you haven't taken this as me calling you out or picking a fight, Because it not that at all. Its more of a two sides of a coin type thing,

Nothing offends me, certainly not rational discussion. I would say its not a two sides of a coin thing, though, Nash Equilibria either exist in non zero sum games or they don't.

It should be noted that these are the longest posts I've ever made on this forum without any sarcasm. . .

Originally Posted by Educ8d Degener8

Holy f*ck you dorks are gonna scare away all the posters from Vaughn's thread with this math/logic circle jerk... ;p

You know you're dying to jump in on this, but you're afraid of being called an aspy.

I think Proposition Joe's bookie wacked him for not paying up on time.

Originally Posted by NunyaBidness

You know you're dying to jump in on this, but you're afraid of being called an aspy.

LOL... Nun - do you f*ck around in any financial market sh*t?

Originally Posted by Educ8d Degener8

LOL... Nun - do you f*ck around in any financial market sh*t?

Nothing serious, some longterm positions. Because of the higher edges and smaller liquidity in sportsbetting compared to traditional investments I think there's a tipping point with bankroll size where it makes more sense to do one than the other. I'm not sure where that number is, but I know I'm nowhere near it.

It would be a lot of work to make the changeover, I'm still learning this game. And I'm even less interested in business than I am sports.

Originally Posted by NunyaBidness

Nothing serious, some longterm positions. Because of the higher edges and smaller liquidity in sportsbetting compared to traditional investments I think there's a tipping point with bankroll size where it makes more sense to do one than the other. I'm not sure where that number is, but I know I'm nowhere near it.

It would be a lot of work to make the changeover, I'm still learning this game. And I'm even less interested in business than I am sports.

I'm gonna send u my NFL picks this season. Three huge seasons in a row!! I even had the Giants winning the Superbowl at +2500 before the playoffs started. To be fair, I also had the same bet on the 49ers at +2500. I felt one of them would win the Superbowl, they ended up facing each other (one game prior to the Super Bowl,) and one of them did win.